The influence of sleep quality, sleep duration and sleepiness on school performance in children and adolescents: A meta-analytic review

k=16 studies / N=13,631 for sleep quality; k=17 / N=15,199 for sleep duration; k=17 / N=19,530 for sleepiness

Children and adolescents in school, pooled across studies

Three separate random-effects meta-analyses with moderator analyses for gender, age, and parameter assessment. The pooled studies are overwhelmingly cross-sectional correlational designs — the meta inherits that confounding entirely, and the authors themselves close by recommending longitudinal and experimental work "in order to gain more insight". Graded C as a meta-analysis of predominantly D-grade primary designs; it is included here not as causal evidence but as the ceiling estimate for how large the raw sleep-achievement correlation is.

All three sleep variables are significantly but MODESTLY related to school performance. Sleepiness shows the strongest relation (r = -0.133), then sleep quality (r = 0.096), then sleep duration (r = 0.069). An r of 0.069 is roughly 0.5% of variance in school performance — and that is the *uncorrected, confounded* correlation, before any adjustment for family background or heritable ability. Effect sizes were larger for younger participants; for sleep duration, age effects were larger in samples with more boys. Anyone arguing that a 30-45 minute sleep gain should translate into a large achievement gain has to explain why the confounded correlation is this small to begin with.

High, and unaddressed. Sleep duration is ~69% heritable in adolescence (Kocevska 2021), and the families who produce short-sleeping teenagers also transmit alleles and home environments that affect school performance. Every pooled correlation here is open to the standard passive gene-environment explanation. This is why the number matters mainly as an upper bound: the causal effect can only be smaller than the confounded association, not larger.